Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Predictive Indicators to Identify High-Risk Paediatric Febrile Neutropenia in Paediatric Oncology Patients in a Middle-Income Country

Predictive Indicators to Identify High-Risk Paediatric Febrile Neutropenia in Paediatric Oncology... ABSTRACT Purpose To validate a clinical risk prediction score (Ammann score) to predict adverse events (AEs) in paediatric febrile neutropenia (FN). Patients and methods Patients <16 years of age were enrolled. A risk prediction score (based on haemoglobin ≥ 9 g/dl, white cell count (WCC) < 0.3 G/l, platelet count <50 G/l and chemotherapy more intensive than acute lymphoblastic leukaemia maintenance therapy) was calculated and AEs were documented. Results In total, 100 FN episodes occurred in 52 patients, male:female ratio was 1.8:1 and median age was 56 months. At reassessment, AEs occurred in 18 of 55 (45%) low-risk FN episodes (score < 9) and 21 of 42 (55%) high-risk episodes (score ≥9) (sensitivity 60%, specificity 65%, positive predictive value 53%, negative predictive value 71%). Total WCC and absolute monocyte count (AMC) were significantly associated with AEs. Conclusion This study identified total WCC and AMC as significantly associated with AEs but failed to validate the risk prediction score. prediction score, febrile neutropenia, childhood malignancy INTRODUCTION Childhood cancer represents 1–10% of all cancers, [1] with an annual incidence of 70–160 per million globally [2] and 45 per million in South Africa [3]. It remains the second most common cause of death in the USA [4] and the UK [5], contributing up to 8% of the post neonatal mortality rate worldwide [6]. Despite improvements in survival because of improved supportive care, febrile neutropenia (FN) remains one of the most common complications of chemotherapy [7, 8]. In FN patients, there is no evidence-based method to rule out an infectious cause of fever. Therefore, FN episodes are managed according to a standard protocol including hospital admission and intravenous antibiotics [9, 10]. Recent guidelines recommend the use of a validated scoring system to assess the risk of developing complications and to individualize patient management [9]. Despite the development of several paediatric risk assessment scores [11–21], none have been internationally validated and none validated in low- to middle-income countries with a high burden of infectious diseases [8, 22]. This prospective study aimed to validate a risk assessment score published by Ammann et al. [21] to distinguish between high- and low-risk patients, who might develop adverse events (AEs) during FN. The Ammann score is based on a weighted score derived from four variables [haemoglobin (Hb) ≥9 g/dl (weight = 5), white cell count (WCC) <0.3 G/L (weight = 3), platelet count <50 G/L (weight =3) and chemotherapy more intensive than acute lymphoblastic leukaemia maintenance therapy (weight = 4)]. A score of ≥ 9 indicates that a patient is at high risk of developing an AE with a 92% sensitivity [21]. The inclusion of an Hb ≥ 9 g/dl was described by Ammann et al. [21] as “seemingly counter intuitive, but it might reflect clinically important dehydration or the previous need for transfusion of packed cells”. A risk prediction model would be of value in a middle-income country with resource limitations and a high burden of infectious diseases [23, 24]. Identification of a low-risk group would enable early step-down from intravenous to oral antibiotics [25] that could benefit the institution financially and reduce patient discomfort. Identification of patients at high risk of an AE would enable intensive monitoring and early recognition of complications. PATIENTS AND METHODS All patients <16 years of age, receiving chemotherapy according to standard treatment protocols (Supplementary Appendix A) at a tertiary-level paediatric oncology unit in Cape Town, South Africa, who presented with fever and neutropenia from 22 January 2014 to 22 January 2016, were prospectively enrolled in the study. Multiple episodes of FN per patient were allowed. On presentation, a clinical assessment, full blood count, differential WCC, serum C-reactive protein (CRP), appropriate cultures and radiological examinations were performed based on the patients’ clinical condition. All patients were managed according to the current standard of care protocol for FN and reassessed by the hospital paediatric oncology team within 24 h of admission. An Ammann score [21] was calculated for each FN episode, and AEs were documented until antibiotics had been stopped or neutropenia resolved. A score of ≥ 9 was considered high risk of developing an AE and a score of < 9 was considered as low risk. Definitions Fever was defined as recorded axillary temperature ≥38.0 °C on ≥ 2 occasions/24 h or ≥38.5 °C once. Neutropenia was defined as an absolute neutrophil count ≤0.5 G/L [21]. AE included severe medical complications (SMCs) as a result of infection (including death, complications requiring intensive care admission and acute life-threatening events), microbiologically defined infection (MDI—positive bacterial/fungal culture from a normally sterile body compartment or the detection of a viral antigen or product of polymerase chain reaction by a validated microbiological method) and radiologically confirmed pneumonia (RCP—the presence of clinical symptoms and radiographic changes suggestive of a pneumonia as reported by a paediatric pulmonologist). Statistical analysis One-way analysis of variance was used to test for differences in means of continuous measurements between the AE and no AE groups. Summary statistics were reported as medians and interquartile ranges. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated from cross-tabulations between the AEs and non-AE groups and the risk score groups. A significance level was set at p < 0.05. The performance of the risk assessment tool at reassessment was calculated according to the calculation used by Ammann et al. (sensitivitycombined = proportionknown + sensitivitypredicted * (1−proportionknown) [21]. The calculation was not applied at presentation as none of the episodes were excluded. RESULTS Epidemiology and description of FN episodes Within the 2-year study period, 100 episodes of FN were reported in 52 patients [median 2 episodes (range 1–5); median patient age 56 months (range 8–175 months), male:female ratio (1.8:1). In total, >18% FN episodes were enrolled in the second year of the study, probably because of the increased number of newly diagnosed childhood malignancies (2015 n = 73; 2014 n = 56). The study population included 28 haematological malignancies (54% FN episodes) and 24 solid tumours (46% FN episodes) (Table 1). Table 1 FN, the malignancies being treated, the stage of disease and number of patients with a central venous catheter Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) CVC, central venous catheter; ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia.. Table 1 FN, the malignancies being treated, the stage of disease and number of patients with a central venous catheter Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) CVC, central venous catheter; ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia.. In total, 24% FN episodes occurred in patients with advanced stage disease (Table 1) and 18% had ≥1 underlying comorbidity (Table 2). A clinical site of infection at presentation could not be identified in 37% of episodes. The most common clinical demonstrable sites of infection were mucositis (15%), acute gastroenteritis/typhlitis (12%) and upper respiratory tract infections (11%) (Table 3). Table 2 Comorbidities associated with FN Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 AML, acute myeloblastic leukaemia; ALL, acute lymphoblastic leukaemia; TB, tuberculosis; HIV, human immunodeficiency virus; DVT, deep vein thrombosis; VRE, vancomycin-resistant colonization. Table 2 Comorbidities associated with FN Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 AML, acute myeloblastic leukaemia; ALL, acute lymphoblastic leukaemia; TB, tuberculosis; HIV, human immunodeficiency virus; DVT, deep vein thrombosis; VRE, vancomycin-resistant colonization. Table 3 Clinical site of infection with associated adverse events Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 URTI, upper respiratory tract infection; LRTI, lower respiratory tract infection; AGE, acute gastroenteritis; CVC, central venous catheter; UTI, urinary tract infection., Table 3 Clinical site of infection with associated adverse events Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 URTI, upper respiratory tract infection; LRTI, lower respiratory tract infection; AGE, acute gastroenteritis; CVC, central venous catheter; UTI, urinary tract infection., Adverse events In 40 of the 100 episodes of FN, ≥1 AE occurred [33 MDIs, 10 RCPs and 9 SMCs (2 ALTE; 4 ICU admissions; 3 deaths)]. Microbiologically defined infections MDIs included 24 bacteraemias, three respiratory tract infections, four urinary tract infections and two skin infections. The majority (58%) were gram-positive bacteraemias [Staphylococcus aureus (n = 4), Streptococcus species (n = 9)] with 41% gram-negative bacteraemia [Klebsiella pneumoniae (n = 6), Escherichia coli (n = 5)]. Polymicrobial bacteraemia detected in four FN episodes. Radiologically confirmed pneumonia RCPs occurred in 10 FN episodes. Twenty-nine chest radiographs were reported as abnormal. Performance of the risk assessment tool AEs were present in three FN episodes at presentation (two low risk, one high risk) and two FN episodes at reassessment (two high risk). By applying the risk assessment score and calculation [21] at reassessment, AEs were correctly predicted in 21 of 42 episodes in the high-risk group and 37 of 55 episodes in the low-risk group, thus yielding a 60% sensitivity, 65% specificity, 53% PPV and 71% NPV. When applying the score at presentation, a 51% sensitivity, 68% specificity, 55% PPV and 65% NPV were calculated (Table 4). Table 4 Performance of the risk prediction score applied to the three different cohorts using the ammann rule At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 a Sensitivity at reassessment calculated according to the calculation used by Ammann et al. b Sensitivity calculated at presentation according to standard methods. Table 4 Performance of the risk prediction score applied to the three different cohorts using the ammann rule At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 a Sensitivity at reassessment calculated according to the calculation used by Ammann et al. b Sensitivity calculated at presentation according to standard methods. We applied the data to risk prediction scores published by other authors [11–13, 16, 17], but none of these scores achieved the predefined sensitivity of ≥ 90% [21] (Table 5). Table 5 Performance of risk prediction scores using additional models as applied to our cohort Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 TP, true positive; FP, false positive; FN, false negative; TN, true negative. Table 5 Performance of risk prediction scores using additional models as applied to our cohort Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 TP, true positive; FP, false positive; FN, false negative; TN, true negative. In a univariate analysis of the predictive performance of individual variables, total WCC at 0.5 G/L (p = 0.01) and absolute monocyte count (AMC) at 0.08 G/L (p = 0.05) achieved statistical significance, while temperature (p = 0.02) and body mass index (p = 0.03) predict an AE in multivariate analysis performed on patients without an AE known at presentation. Serum CRP in the group that developed an AE did not differ from those who did not develop and AE (p = 0.42) (Table 6). Table 6 Performance of individual variables in predicting adverse events AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 *Statistically significant. BMI, body mass index; G/l = cells ×109/l. Table 6 Performance of individual variables in predicting adverse events AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 *Statistically significant. BMI, body mass index; G/l = cells ×109/l. DISCUSSION This study aimed to validate a risk assessment score published by Ammann et al. [21] that could lead to the identification of patients at high risk of suffering an AE during a FN episode. In the derivation and validation of Ammann's score [21], the predefined sensitivity was established at ≥90%. When the score was applied at reassessment in Ammann's cohort, it achieved a 92% sensitivity, 45% specificity, 93% NPV and 40% PPV [21]. Miedema et al. could not reach these numbers in a Dutch cohort (82% sensitivity, 57%specificity, 23% PPV, 91% NPV). He proposed differences in chemotherapy treatment protocols, genetic factors, microbiological environments and retrospective data collection, as reasons for the lower sensitivity. In this study, the score was applied prospectively, but yielded a 60% sensitivity, 65% specificity, 53% PPV and 71% NPV at reassessment. By applying Amman’s calculation used in both of the other study models, the sensitivity increased at reassessment. This calculation however assumes that the excluded cases were correctly predicted. In our cohort, two of the three excluded cases were incorrectly predicted, resulting in a false increase in sensitivity at reassessment. When the score was applied at presentation, the model yielded a 51% sensitivity, 68% specificity, 55% PPV and 65% NPV. Miedema et al. [26] demonstrated a similar performance of the prediction score at presentation (69% sensitivity, 57% specificity, 34% PPV, 85% NPV). One possible explanation for the difference in the performance of the prediction score might be differences between the European and South African cohorts. Our cohort included younger patients (4.7 vs. 6.9 years) [21], more patients with solid tumours (44 vs. 26%) [21], a higher AE rate (40 vs. 29%) [21], more bacteraemias (24 vs. 15.8%) [21] and SMCs (9 vs. 4.9%) [21]. We reported a 30% bacteraemia rate in the high-risk population vs. 7% by Ammann et al. [21] and 29% by Miedema et al. [26] Thus, 46% of bacteraemia occurred in patients predicted as low risk, including one episode that resulted in death. Miedema et al. [26] found that almost one in every three patients (29%) with bacteraemia were incorrectly classified and in the Ammann [21] cohort two of three deaths occurred in low-risk episodes. The low sensitivity in two of the three cohorts and the high number of incorrectly classified episodes in all three cohorts limit the applicability of this prediction score outside of the derivation population. Another explanatory factor may be the higher number of patients assessed at low risk for developing an AE (57 vs. 35%). In calculating the score, an Hb ≥ 9 g/dl is included as the highest weighted variable. In our cohort, the mean Hb level was 8.1 g/dl on presentation, thus contributing to a lower-risk prediction score and a higher number of low-risk patients. The lower mean Hb might be explained by the adherence to international guidelines reserving blood transfusions for patients with an Hb < 7 g/dl [27]. Many of the FN episodes occurred in patients with advanced stage of disease (24%), relapsed malignancies (16%) and solid tumours outside the central nervous system (CNS) (46%), resulting in the use of more aggressive chemotherapy protocols causing more severe bone marrow suppression and more AEs. This supports Stones et al.'s [28] finding that advanced stage disease contributes to a poorer outcome and higher incidence of AEs. This is compounded by the restricted use of granulocyte colony-stimulating factor, which is reserved for cases of severe prolonged neutropenia. Management decisions are often based on laboratory parameters including serum CRP values and differential WCC. As demonstrated in this study and confirmed by other studies, CRP is of limited value [29–32]. Total WCC and AMC achieved statistical significance in univariate analysis, however their applicability in predicting AEs where limited by the low sensitivity and specificity as individual predictors (Table 6). Although AMC has been used in previously defined risk prediction scores, these could not be validated [11, 12, 14], most likely because statistical significance was reached at a lower AMC level (AMC < 0.08 G/l) than in other studies [11, 12] Another limitation was the inclusion and exclusion criteria, which could not be applied strictly, as there is variability in the definitions used by the different authors [29–32]. In conclusion, although this study did not succeed in validating the Ammann score [21], it demonstrated the important significant association between total WCC and AMC in an AE during FN. We demonstrate that other prediction scores [11–13, 16, 17] are not useful in our patient population and that a risk prediction score should be developed for low- to middle-income countries with a high burden of infectious diseases. Finally, the study demonstrated the marked differences in patient cohorts between high-income countries vs. a low- to middle-income country. Future studies should include the development of a cost-effective model, including Procalcitonin (PCT) to assist with differentiating between high-risk and low-risk FN, as PCT has been demonstrated to be a more sensitive marker for bacterial infection [29–31]. ACKNOWLEDGEMENT The authors wish to acknowledge Professor Robert Gie for his input and support in preparing the manuscript. REFERENCES 1 Ferlay J , Bray F , Pisani P , et al. GLOBOCAN 2002: cancer incidence, mortality and prevalence worldwide. IARC Cancer Base No. 5, Version 2.0. Lyon : International Agency for Research on Cancer Press , 2004 . 2 Parkin DM , Kremárová E , Draper GJ , et al. International Incidence of Childhood Cancer . Vol. 2 . Lyon, France: IARC Scientific Publication , WHO Press, 1998 . 3 Stefan DC , Stones DK , Wainwright RD , et al. Childhood cancer incidence in South Africa, 1987–2007 . S Afr Med J 2015 ; 105 : 939 – 47 . Google Scholar Crossref Search ADS PubMed 4 Ten leading causes of death by age group, United States—2013 [Online]. 2014 . http://www.cdc.gov/injury/images/lc_charts (15 April 2016, date last accessed). 5 Wolfe I , Macfarlane A , Donkin A , et al. Why children die: death in infants, children and young people in the UK, Part A. London, UK: Royal College of Paediatric and Child Health and National Children’s Bureau, 2014 . 6 Child causes of death 2000–2015 . WHO global health estimates technical paper—WHO/HIC/IER/GHE/2016.1 [Online]. 2015 . www.who.int/gho/child_health/mortality/causes (15 April 2016, date last accessed). 7 Freycon F , Trombert-Paviot B. Trends in treatment-related deaths (TRDs) in childhood cancer and leukemia over time: a follow-up of patients included in the childhood cancer registry of the Rh‘‘ne-Alpes region in France (ARCERRA) . Pediatr Blood Cancer 2008 ; 50 : 1213 – 20 . Google Scholar Crossref Search ADS PubMed 8 Phillips B , Wade R , Stewart LA , et al. Systematic review and meta-analysis of the discriminatory performance of risk prediction rules in febrile neutropaenic episodes in children and young people . Eur J Cancer 46 : 2950 – 64 . Crossref Search ADS PubMed 9 Bate J , Gibson F , Johnson E , et al. Neutropenic sepsis: prevention and management of neutropenic sepsis in cancer patients (NICE Clinical Guideline CG151) . Arch Dis Child Educ Pract Ed 2013 ; 98 : 73 – 5 . Google Scholar Crossref Search ADS PubMed 10 Henry M , Sung L. Supportive care in pediatric oncology: oncologic emergencies and management of fever and neutropenia . Pediatr Clin North Am 2015 ; 62 : 27 – 6 . Google Scholar Crossref Search ADS PubMed 11 Rackoff WR , Gonin R , Robinson C , et al. Predicting the risk of bacteremia in children with fever and neutropenia . J Clin Oncol 1996 ; 14 : 919 – 24 . Google Scholar Crossref Search ADS PubMed 12 Baorto PB , Aquino VM , Mullen CA , et al. Clinical parameters associated with low bacteremia risk in 1 100 pediatric oncology patients with fever and neutropenia . Am Cancer Soc 2001 ; 92 : 909 – 13 . 13 Rondinelli PI , Ribeiro Kde C , De Camargo B. A proposed score for predicting severe infection complications in children with chemotherapy-induced febrile neutropenia . J Pediatr Hematol Oncol 2006 ; 28 : 665 – 70 . Google Scholar Crossref Search ADS PubMed 14 Alexander SW , Wade KC , Hibberd PL , et al. Evaluation of risk prediction criteria for episodes of febrile neutropenia in children with cancer . J Pediatr Hematol Oncol 24 : 38 – 42 . Crossref Search ADS PubMed 15 Klaassen RJ , Goodman TR , Pham B , et al. “Low-risk” prediction rule for pediatric oncology patients presenting with fever and neutropenia . J Clin Oncol 2000 ; 18 : 1012 – 19 . Google Scholar Crossref Search ADS PubMed 16 Santolaya ME , Alvarez AM , Becker A , et al. Prospective, multicentre evaluation of risk factors associated with invasive bacterial infection in children with cancer, neutropenia, and fever . J Clin Oncol 2001 ; 19 : 3415 – 21 . Google Scholar Crossref Search ADS PubMed 17 Santolaya ME , Alvarez AM , Avilés CL , et al. Prospective evaluation of a model of prediction of invasive bacterial infection risk among children with cancer, fever, and neutropenia . Clin Infect Dis 2002 ; 35 : 678 – 83 . Google Scholar Crossref Search ADS PubMed 18 Paganini H , Gomez S , Ruvinsky S , et al. Outpatient, sequential, parenteral-oral antibiotic therapy for lower risk febrile neutropenia in children with malignant disease: a single-centre, randomized, controlled trial in Argentina . Cancer 2003 ; 97 : 1775 – 80 . Google Scholar Crossref Search ADS PubMed 19 Ammann RA , Hirt A , Lüthy AR , et al. Identification of children presenting with fever in chemotherapy-induced neutropenia at low risk for severe bacterial infection . Med Pediatr Oncol 41 : 436 – 43 . Crossref Search ADS PubMed 20 Agyeman P , Aebi C , Hirt A , et al. Predicting bacteremia in children with cancer and fever in chemotherapy-induced neutropenia: results of the prospective multicenter SPOG 2003 FN study . Pediatr Infect Dis J 2011 ; 30 : e114 – 19 . Google Scholar Crossref Search ADS PubMed 21 Ammann RA , Bodmer N , Hirt A , et al. Predicting adverse events in children with fever and chemo-therapy-induced neutropenia: the prospective multicenter SPOG 2003 FN study . J Clin Oncol 2010 ; 28 : 2008 – 14 . Google Scholar Crossref Search ADS PubMed 22 Phillips RS , Lehrnbecher T , Alexander S , et al. Updated systematic review and meta-analysis of the performance of risk prediction rules in children and young people with febrile neutropenia . PLoS One 2013 ; 7 : e38300 . [Online] 2012 [access July 4]; Available: www.plosone.org. Google Scholar Crossref Search ADS 23 Brack E , Bodmer N , Simon A , et al. First-day step-down to oral outpatient treatment versus continued standard treatment in children with cancer and low-risk fever in neutropenia. A randomized controlled trial within the multicenter SPOG 2003 FN Study . Pediatr Blood Cancer 2012 ; 59 : 423 – 30 . Google Scholar Crossref Search ADS PubMed 24 WHO . Global tuberculosis control 2012. Geneva: World Health Organization, 2012 . 25 Dodd PJ , Gardiner E , Coghlan R , et al. Burden of childhood tuberculosis in 22 high-burden countries: a mathematical modelling study . Lancet Glob Health 2014 ; 2 : 453 – 9 . Google Scholar Crossref Search ADS 26 Miedema KG , De Bont ES , Oude Nijhuis CS , et al. Validation of a new risk assessment model for predicting adverse events in children with fever and chemotherapy-induced neutropenia . J Clin Oncol 2011 ; 29 : e182 – 4 . Google Scholar Crossref Search ADS PubMed 27 Gomez E , Agrawal AK , Hastings CA , Transfusion support. In: Supportive care in Pediatric Oncology: A Practical Evidence-based Approach, Pediatric Oncology . Berlin . Springer Heidelberg , 2015 . 28 Stones DK , De Bruin GP , Esterhuizen TM , et al. Childhood cancer survival rates in two South African units . S Afr Med J 2014 ; 104 : 502 – 4 . Google Scholar Crossref Search ADS 29 Secmeer G , Devrim I , Kara A , et al. Role of procalcitonin and CRP in differentiating a stable from a deteriorating clinical course in pediatric febrile neutropenia . J Pediatr Hematol Oncol 2007 ; 29 : 107 – 11 . Google Scholar Crossref Search ADS PubMed 30 Miedema KG , De Bont ES , Elferink RF , et al. The diagnostic value of CRP, IL-8, PCT, and sTREM-1 in the detection of bacterial infections in pediatric oncology patients with febrile neutropenia . Support Care Cancer 2011 ; 19 : 1593 – 600 . Google Scholar Crossref Search ADS PubMed 31 Kitanovski L , Jazbec J , Hojker S , et al. Diagnostic accuracy of procalcitonin and interleukin-6 values for predicting bacteremia and clinical sepsis in febrile neutropenic children with cancer . Eur J Clin Microbiol Infect Dis 2006 ; 25 : 413 – 15 . Google Scholar Crossref Search ADS PubMed 32 Mian A , Becton D , Saylors R , et al. Biomarkers for risk stratification of febrile neutropenia among children with malignancy: a pilot study . Pediatr Blood Cancer 2012 ; 59 : 238 – 45 . Google Scholar Crossref Search ADS PubMed © The Author [2017]. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Tropical Pediatrics Oxford University Press

Predictive Indicators to Identify High-Risk Paediatric Febrile Neutropenia in Paediatric Oncology Patients in a Middle-Income Country

Download PDF
8 pages

Loading next page...
 
/lp/ou_press/predictive-indicators-to-identify-high-risk-paediatric-febrile-70HC0LmRvh

References (38)

Publisher
Oxford University Press
Copyright
© The Author [2017]. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0142-6338
eISSN
1465-3664
DOI
10.1093/tropej/fmx082
Publisher site
See Article on Publisher Site

Abstract

ABSTRACT Purpose To validate a clinical risk prediction score (Ammann score) to predict adverse events (AEs) in paediatric febrile neutropenia (FN). Patients and methods Patients <16 years of age were enrolled. A risk prediction score (based on haemoglobin ≥ 9 g/dl, white cell count (WCC) < 0.3 G/l, platelet count <50 G/l and chemotherapy more intensive than acute lymphoblastic leukaemia maintenance therapy) was calculated and AEs were documented. Results In total, 100 FN episodes occurred in 52 patients, male:female ratio was 1.8:1 and median age was 56 months. At reassessment, AEs occurred in 18 of 55 (45%) low-risk FN episodes (score < 9) and 21 of 42 (55%) high-risk episodes (score ≥9) (sensitivity 60%, specificity 65%, positive predictive value 53%, negative predictive value 71%). Total WCC and absolute monocyte count (AMC) were significantly associated with AEs. Conclusion This study identified total WCC and AMC as significantly associated with AEs but failed to validate the risk prediction score. prediction score, febrile neutropenia, childhood malignancy INTRODUCTION Childhood cancer represents 1–10% of all cancers, [1] with an annual incidence of 70–160 per million globally [2] and 45 per million in South Africa [3]. It remains the second most common cause of death in the USA [4] and the UK [5], contributing up to 8% of the post neonatal mortality rate worldwide [6]. Despite improvements in survival because of improved supportive care, febrile neutropenia (FN) remains one of the most common complications of chemotherapy [7, 8]. In FN patients, there is no evidence-based method to rule out an infectious cause of fever. Therefore, FN episodes are managed according to a standard protocol including hospital admission and intravenous antibiotics [9, 10]. Recent guidelines recommend the use of a validated scoring system to assess the risk of developing complications and to individualize patient management [9]. Despite the development of several paediatric risk assessment scores [11–21], none have been internationally validated and none validated in low- to middle-income countries with a high burden of infectious diseases [8, 22]. This prospective study aimed to validate a risk assessment score published by Ammann et al. [21] to distinguish between high- and low-risk patients, who might develop adverse events (AEs) during FN. The Ammann score is based on a weighted score derived from four variables [haemoglobin (Hb) ≥9 g/dl (weight = 5), white cell count (WCC) <0.3 G/L (weight = 3), platelet count <50 G/L (weight =3) and chemotherapy more intensive than acute lymphoblastic leukaemia maintenance therapy (weight = 4)]. A score of ≥ 9 indicates that a patient is at high risk of developing an AE with a 92% sensitivity [21]. The inclusion of an Hb ≥ 9 g/dl was described by Ammann et al. [21] as “seemingly counter intuitive, but it might reflect clinically important dehydration or the previous need for transfusion of packed cells”. A risk prediction model would be of value in a middle-income country with resource limitations and a high burden of infectious diseases [23, 24]. Identification of a low-risk group would enable early step-down from intravenous to oral antibiotics [25] that could benefit the institution financially and reduce patient discomfort. Identification of patients at high risk of an AE would enable intensive monitoring and early recognition of complications. PATIENTS AND METHODS All patients <16 years of age, receiving chemotherapy according to standard treatment protocols (Supplementary Appendix A) at a tertiary-level paediatric oncology unit in Cape Town, South Africa, who presented with fever and neutropenia from 22 January 2014 to 22 January 2016, were prospectively enrolled in the study. Multiple episodes of FN per patient were allowed. On presentation, a clinical assessment, full blood count, differential WCC, serum C-reactive protein (CRP), appropriate cultures and radiological examinations were performed based on the patients’ clinical condition. All patients were managed according to the current standard of care protocol for FN and reassessed by the hospital paediatric oncology team within 24 h of admission. An Ammann score [21] was calculated for each FN episode, and AEs were documented until antibiotics had been stopped or neutropenia resolved. A score of ≥ 9 was considered high risk of developing an AE and a score of < 9 was considered as low risk. Definitions Fever was defined as recorded axillary temperature ≥38.0 °C on ≥ 2 occasions/24 h or ≥38.5 °C once. Neutropenia was defined as an absolute neutrophil count ≤0.5 G/L [21]. AE included severe medical complications (SMCs) as a result of infection (including death, complications requiring intensive care admission and acute life-threatening events), microbiologically defined infection (MDI—positive bacterial/fungal culture from a normally sterile body compartment or the detection of a viral antigen or product of polymerase chain reaction by a validated microbiological method) and radiologically confirmed pneumonia (RCP—the presence of clinical symptoms and radiographic changes suggestive of a pneumonia as reported by a paediatric pulmonologist). Statistical analysis One-way analysis of variance was used to test for differences in means of continuous measurements between the AE and no AE groups. Summary statistics were reported as medians and interquartile ranges. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated from cross-tabulations between the AEs and non-AE groups and the risk score groups. A significance level was set at p < 0.05. The performance of the risk assessment tool at reassessment was calculated according to the calculation used by Ammann et al. (sensitivitycombined = proportionknown + sensitivitypredicted * (1−proportionknown) [21]. The calculation was not applied at presentation as none of the episodes were excluded. RESULTS Epidemiology and description of FN episodes Within the 2-year study period, 100 episodes of FN were reported in 52 patients [median 2 episodes (range 1–5); median patient age 56 months (range 8–175 months), male:female ratio (1.8:1). In total, >18% FN episodes were enrolled in the second year of the study, probably because of the increased number of newly diagnosed childhood malignancies (2015 n = 73; 2014 n = 56). The study population included 28 haematological malignancies (54% FN episodes) and 24 solid tumours (46% FN episodes) (Table 1). Table 1 FN, the malignancies being treated, the stage of disease and number of patients with a central venous catheter Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) CVC, central venous catheter; ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia.. Table 1 FN, the malignancies being treated, the stage of disease and number of patients with a central venous catheter Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) Malignancy n = patients (% of FN) Relapsed malignancy n = patients (% of FN) Advanced stage n = patients (% of FN) CVC n = patients (% of FN) ALL 15 (29) 2 (9) 7 (9) AML 9 (24) 6 (12) Lymphoma 4 (5) Rhabdomyosarcoma 4 (6) 4 (6) 1 (1) Ewing sarcoma 4 (6) 1 (1) Retinoblastoma 3 (8) 1 (6) 1 (6) Nephroblastoma 3 (7) 1 (3) Osteosarcoma 3 (3) 1 (1) 2 (2) Neuroblastoma 3 (7) 3 (7) 1 (1) CNS tumours 2 (2) Other 2 (3) Total 52 (100) 4 (16) 11 (24) 16 (24) CVC, central venous catheter; ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia.. In total, 24% FN episodes occurred in patients with advanced stage disease (Table 1) and 18% had ≥1 underlying comorbidity (Table 2). A clinical site of infection at presentation could not be identified in 37% of episodes. The most common clinical demonstrable sites of infection were mucositis (15%), acute gastroenteritis/typhlitis (12%) and upper respiratory tract infections (11%) (Table 3). Table 2 Comorbidities associated with FN Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 AML, acute myeloblastic leukaemia; ALL, acute lymphoblastic leukaemia; TB, tuberculosis; HIV, human immunodeficiency virus; DVT, deep vein thrombosis; VRE, vancomycin-resistant colonization. Table 2 Comorbidities associated with FN Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 Comorbidity Malignancy Number of FN episodes AE Trisomy 21 AML 3 1 Trisomy 21 AML 1 1 Trisomy 21 Pre-B-cell ALL 2 1 Dilated cardiomyopathy Nephroblastoma 2 1 Dilated cardiomyopathy Ewing sarcoma 1 None Primary variable immunodeficiency B-cell ALL 1 1 Hypertension Neuroblastoma 2 1 Drug-induced tubulopathy T-cell ALL 1 None Pulmonary TB Osteosarcoma 1 None HIV Kaposi sarcoma 1 None Femoral DVT Pre-B-cell ALL 1 None VRE colonization AML 2 1 Total 18 7 AML, acute myeloblastic leukaemia; ALL, acute lymphoblastic leukaemia; TB, tuberculosis; HIV, human immunodeficiency virus; DVT, deep vein thrombosis; VRE, vancomycin-resistant colonization. Table 3 Clinical site of infection with associated adverse events Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 URTI, upper respiratory tract infection; LRTI, lower respiratory tract infection; AGE, acute gastroenteritis; CVC, central venous catheter; UTI, urinary tract infection., Table 3 Clinical site of infection with associated adverse events Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 Clinical site Outcome Type of adverse event Total AE No AE MDI SMC RCP Fever of unknown origin 37 10 27 8 2 2 URTI 11 3 8 2 1 0 LRTI 4 3 1 1 1 3 AGE/typhlitis 12 2 10 1 1 1 Mucositis 15 7 9 6 1 0 Skin/soft tissue infection 4 3 1 2 1 1 CVC-related infection 3 2 1 2 – – UTI 3 3 0 3 – – Other 5 3 2 3 – – ≥1 site of infection 6 4 2 2 1 3 URTI, upper respiratory tract infection; LRTI, lower respiratory tract infection; AGE, acute gastroenteritis; CVC, central venous catheter; UTI, urinary tract infection., Adverse events In 40 of the 100 episodes of FN, ≥1 AE occurred [33 MDIs, 10 RCPs and 9 SMCs (2 ALTE; 4 ICU admissions; 3 deaths)]. Microbiologically defined infections MDIs included 24 bacteraemias, three respiratory tract infections, four urinary tract infections and two skin infections. The majority (58%) were gram-positive bacteraemias [Staphylococcus aureus (n = 4), Streptococcus species (n = 9)] with 41% gram-negative bacteraemia [Klebsiella pneumoniae (n = 6), Escherichia coli (n = 5)]. Polymicrobial bacteraemia detected in four FN episodes. Radiologically confirmed pneumonia RCPs occurred in 10 FN episodes. Twenty-nine chest radiographs were reported as abnormal. Performance of the risk assessment tool AEs were present in three FN episodes at presentation (two low risk, one high risk) and two FN episodes at reassessment (two high risk). By applying the risk assessment score and calculation [21] at reassessment, AEs were correctly predicted in 21 of 42 episodes in the high-risk group and 37 of 55 episodes in the low-risk group, thus yielding a 60% sensitivity, 65% specificity, 53% PPV and 71% NPV. When applying the score at presentation, a 51% sensitivity, 68% specificity, 55% PPV and 65% NPV were calculated (Table 4). Table 4 Performance of the risk prediction score applied to the three different cohorts using the ammann rule At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 a Sensitivity at reassessment calculated according to the calculation used by Ammann et al. b Sensitivity calculated at presentation according to standard methods. Table 4 Performance of the risk prediction score applied to the three different cohorts using the ammann rule At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 At reassessmenta At presentationb Time of assessment Present cohort (%) Ammann et al. (%) Miedema et al. (%) Present cohort (%) Miedema et al. (%) Sensitivity 60 92 82 51 69 Specificity 65 45 57 68 57 PPV 53 40 23 55 34 NPV 71 93 91 65 85 a Sensitivity at reassessment calculated according to the calculation used by Ammann et al. b Sensitivity calculated at presentation according to standard methods. We applied the data to risk prediction scores published by other authors [11–13, 16, 17], but none of these scores achieved the predefined sensitivity of ≥ 90% [21] (Table 5). Table 5 Performance of risk prediction scores using additional models as applied to our cohort Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 TP, true positive; FP, false positive; FN, false negative; TN, true negative. Table 5 Performance of risk prediction scores using additional models as applied to our cohort Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 Risk prediction rule Number of TP No. of FP Number of FN Number of TN Sensitivity Specificity NPV PPV Rules predicting bacteraemia Rackoff et al. [11] 30 40 10 18 0.75 0.31 0.64 0.43 Baorto et al. [12] 30 38 8 18 0.79 0.32 0.69 0.44 Rules predicting severe/invasive bacterial infection Rondinelli et al. [13] 7 5 15 25 0.32 0.83 0.63 0.58 Santolaya et al. [16, 17] 27 41 13 19 0.68 0.37 0.59 0.40 TP, true positive; FP, false positive; FN, false negative; TN, true negative. In a univariate analysis of the predictive performance of individual variables, total WCC at 0.5 G/L (p = 0.01) and absolute monocyte count (AMC) at 0.08 G/L (p = 0.05) achieved statistical significance, while temperature (p = 0.02) and body mass index (p = 0.03) predict an AE in multivariate analysis performed on patients without an AE known at presentation. Serum CRP in the group that developed an AE did not differ from those who did not develop and AE (p = 0.42) (Table 6). Table 6 Performance of individual variables in predicting adverse events AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 *Statistically significant. BMI, body mass index; G/l = cells ×109/l. Table 6 Performance of individual variables in predicting adverse events AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 AE group (n) Non-AE group (n) p-value Age (months)—mean 63 76 0.18 BMI (SD)—mean 1.8 1.9 0.64 Sex (male) 31 45 0.77 Temperature (°C)—mean 38 37.8 0.69 WCC (G/l)—mean 0.53 1.02 <0.01* AMC (G/l)—mean (excluding two episodes with unknown AMC) 0.08 0.23 0.05* Hb (g/dl)—mean 8.2 8.0 0.7 Platelets (G/l)—mean 61 94 0.08 CRP (mg/l) 128 112 0.42 *Statistically significant. BMI, body mass index; G/l = cells ×109/l. DISCUSSION This study aimed to validate a risk assessment score published by Ammann et al. [21] that could lead to the identification of patients at high risk of suffering an AE during a FN episode. In the derivation and validation of Ammann's score [21], the predefined sensitivity was established at ≥90%. When the score was applied at reassessment in Ammann's cohort, it achieved a 92% sensitivity, 45% specificity, 93% NPV and 40% PPV [21]. Miedema et al. could not reach these numbers in a Dutch cohort (82% sensitivity, 57%specificity, 23% PPV, 91% NPV). He proposed differences in chemotherapy treatment protocols, genetic factors, microbiological environments and retrospective data collection, as reasons for the lower sensitivity. In this study, the score was applied prospectively, but yielded a 60% sensitivity, 65% specificity, 53% PPV and 71% NPV at reassessment. By applying Amman’s calculation used in both of the other study models, the sensitivity increased at reassessment. This calculation however assumes that the excluded cases were correctly predicted. In our cohort, two of the three excluded cases were incorrectly predicted, resulting in a false increase in sensitivity at reassessment. When the score was applied at presentation, the model yielded a 51% sensitivity, 68% specificity, 55% PPV and 65% NPV. Miedema et al. [26] demonstrated a similar performance of the prediction score at presentation (69% sensitivity, 57% specificity, 34% PPV, 85% NPV). One possible explanation for the difference in the performance of the prediction score might be differences between the European and South African cohorts. Our cohort included younger patients (4.7 vs. 6.9 years) [21], more patients with solid tumours (44 vs. 26%) [21], a higher AE rate (40 vs. 29%) [21], more bacteraemias (24 vs. 15.8%) [21] and SMCs (9 vs. 4.9%) [21]. We reported a 30% bacteraemia rate in the high-risk population vs. 7% by Ammann et al. [21] and 29% by Miedema et al. [26] Thus, 46% of bacteraemia occurred in patients predicted as low risk, including one episode that resulted in death. Miedema et al. [26] found that almost one in every three patients (29%) with bacteraemia were incorrectly classified and in the Ammann [21] cohort two of three deaths occurred in low-risk episodes. The low sensitivity in two of the three cohorts and the high number of incorrectly classified episodes in all three cohorts limit the applicability of this prediction score outside of the derivation population. Another explanatory factor may be the higher number of patients assessed at low risk for developing an AE (57 vs. 35%). In calculating the score, an Hb ≥ 9 g/dl is included as the highest weighted variable. In our cohort, the mean Hb level was 8.1 g/dl on presentation, thus contributing to a lower-risk prediction score and a higher number of low-risk patients. The lower mean Hb might be explained by the adherence to international guidelines reserving blood transfusions for patients with an Hb < 7 g/dl [27]. Many of the FN episodes occurred in patients with advanced stage of disease (24%), relapsed malignancies (16%) and solid tumours outside the central nervous system (CNS) (46%), resulting in the use of more aggressive chemotherapy protocols causing more severe bone marrow suppression and more AEs. This supports Stones et al.'s [28] finding that advanced stage disease contributes to a poorer outcome and higher incidence of AEs. This is compounded by the restricted use of granulocyte colony-stimulating factor, which is reserved for cases of severe prolonged neutropenia. Management decisions are often based on laboratory parameters including serum CRP values and differential WCC. As demonstrated in this study and confirmed by other studies, CRP is of limited value [29–32]. Total WCC and AMC achieved statistical significance in univariate analysis, however their applicability in predicting AEs where limited by the low sensitivity and specificity as individual predictors (Table 6). Although AMC has been used in previously defined risk prediction scores, these could not be validated [11, 12, 14], most likely because statistical significance was reached at a lower AMC level (AMC < 0.08 G/l) than in other studies [11, 12] Another limitation was the inclusion and exclusion criteria, which could not be applied strictly, as there is variability in the definitions used by the different authors [29–32]. In conclusion, although this study did not succeed in validating the Ammann score [21], it demonstrated the important significant association between total WCC and AMC in an AE during FN. We demonstrate that other prediction scores [11–13, 16, 17] are not useful in our patient population and that a risk prediction score should be developed for low- to middle-income countries with a high burden of infectious diseases. Finally, the study demonstrated the marked differences in patient cohorts between high-income countries vs. a low- to middle-income country. Future studies should include the development of a cost-effective model, including Procalcitonin (PCT) to assist with differentiating between high-risk and low-risk FN, as PCT has been demonstrated to be a more sensitive marker for bacterial infection [29–31]. ACKNOWLEDGEMENT The authors wish to acknowledge Professor Robert Gie for his input and support in preparing the manuscript. REFERENCES 1 Ferlay J , Bray F , Pisani P , et al. GLOBOCAN 2002: cancer incidence, mortality and prevalence worldwide. IARC Cancer Base No. 5, Version 2.0. Lyon : International Agency for Research on Cancer Press , 2004 . 2 Parkin DM , Kremárová E , Draper GJ , et al. International Incidence of Childhood Cancer . Vol. 2 . Lyon, France: IARC Scientific Publication , WHO Press, 1998 . 3 Stefan DC , Stones DK , Wainwright RD , et al. Childhood cancer incidence in South Africa, 1987–2007 . S Afr Med J 2015 ; 105 : 939 – 47 . Google Scholar Crossref Search ADS PubMed 4 Ten leading causes of death by age group, United States—2013 [Online]. 2014 . http://www.cdc.gov/injury/images/lc_charts (15 April 2016, date last accessed). 5 Wolfe I , Macfarlane A , Donkin A , et al. Why children die: death in infants, children and young people in the UK, Part A. London, UK: Royal College of Paediatric and Child Health and National Children’s Bureau, 2014 . 6 Child causes of death 2000–2015 . WHO global health estimates technical paper—WHO/HIC/IER/GHE/2016.1 [Online]. 2015 . www.who.int/gho/child_health/mortality/causes (15 April 2016, date last accessed). 7 Freycon F , Trombert-Paviot B. Trends in treatment-related deaths (TRDs) in childhood cancer and leukemia over time: a follow-up of patients included in the childhood cancer registry of the Rh‘‘ne-Alpes region in France (ARCERRA) . Pediatr Blood Cancer 2008 ; 50 : 1213 – 20 . Google Scholar Crossref Search ADS PubMed 8 Phillips B , Wade R , Stewart LA , et al. Systematic review and meta-analysis of the discriminatory performance of risk prediction rules in febrile neutropaenic episodes in children and young people . Eur J Cancer 46 : 2950 – 64 . Crossref Search ADS PubMed 9 Bate J , Gibson F , Johnson E , et al. Neutropenic sepsis: prevention and management of neutropenic sepsis in cancer patients (NICE Clinical Guideline CG151) . Arch Dis Child Educ Pract Ed 2013 ; 98 : 73 – 5 . Google Scholar Crossref Search ADS PubMed 10 Henry M , Sung L. Supportive care in pediatric oncology: oncologic emergencies and management of fever and neutropenia . Pediatr Clin North Am 2015 ; 62 : 27 – 6 . Google Scholar Crossref Search ADS PubMed 11 Rackoff WR , Gonin R , Robinson C , et al. Predicting the risk of bacteremia in children with fever and neutropenia . J Clin Oncol 1996 ; 14 : 919 – 24 . Google Scholar Crossref Search ADS PubMed 12 Baorto PB , Aquino VM , Mullen CA , et al. Clinical parameters associated with low bacteremia risk in 1 100 pediatric oncology patients with fever and neutropenia . Am Cancer Soc 2001 ; 92 : 909 – 13 . 13 Rondinelli PI , Ribeiro Kde C , De Camargo B. A proposed score for predicting severe infection complications in children with chemotherapy-induced febrile neutropenia . J Pediatr Hematol Oncol 2006 ; 28 : 665 – 70 . Google Scholar Crossref Search ADS PubMed 14 Alexander SW , Wade KC , Hibberd PL , et al. Evaluation of risk prediction criteria for episodes of febrile neutropenia in children with cancer . J Pediatr Hematol Oncol 24 : 38 – 42 . Crossref Search ADS PubMed 15 Klaassen RJ , Goodman TR , Pham B , et al. “Low-risk” prediction rule for pediatric oncology patients presenting with fever and neutropenia . J Clin Oncol 2000 ; 18 : 1012 – 19 . Google Scholar Crossref Search ADS PubMed 16 Santolaya ME , Alvarez AM , Becker A , et al. Prospective, multicentre evaluation of risk factors associated with invasive bacterial infection in children with cancer, neutropenia, and fever . J Clin Oncol 2001 ; 19 : 3415 – 21 . Google Scholar Crossref Search ADS PubMed 17 Santolaya ME , Alvarez AM , Avilés CL , et al. Prospective evaluation of a model of prediction of invasive bacterial infection risk among children with cancer, fever, and neutropenia . Clin Infect Dis 2002 ; 35 : 678 – 83 . Google Scholar Crossref Search ADS PubMed 18 Paganini H , Gomez S , Ruvinsky S , et al. Outpatient, sequential, parenteral-oral antibiotic therapy for lower risk febrile neutropenia in children with malignant disease: a single-centre, randomized, controlled trial in Argentina . Cancer 2003 ; 97 : 1775 – 80 . Google Scholar Crossref Search ADS PubMed 19 Ammann RA , Hirt A , Lüthy AR , et al. Identification of children presenting with fever in chemotherapy-induced neutropenia at low risk for severe bacterial infection . Med Pediatr Oncol 41 : 436 – 43 . Crossref Search ADS PubMed 20 Agyeman P , Aebi C , Hirt A , et al. Predicting bacteremia in children with cancer and fever in chemotherapy-induced neutropenia: results of the prospective multicenter SPOG 2003 FN study . Pediatr Infect Dis J 2011 ; 30 : e114 – 19 . Google Scholar Crossref Search ADS PubMed 21 Ammann RA , Bodmer N , Hirt A , et al. Predicting adverse events in children with fever and chemo-therapy-induced neutropenia: the prospective multicenter SPOG 2003 FN study . J Clin Oncol 2010 ; 28 : 2008 – 14 . Google Scholar Crossref Search ADS PubMed 22 Phillips RS , Lehrnbecher T , Alexander S , et al. Updated systematic review and meta-analysis of the performance of risk prediction rules in children and young people with febrile neutropenia . PLoS One 2013 ; 7 : e38300 . [Online] 2012 [access July 4]; Available: www.plosone.org. Google Scholar Crossref Search ADS 23 Brack E , Bodmer N , Simon A , et al. First-day step-down to oral outpatient treatment versus continued standard treatment in children with cancer and low-risk fever in neutropenia. A randomized controlled trial within the multicenter SPOG 2003 FN Study . Pediatr Blood Cancer 2012 ; 59 : 423 – 30 . Google Scholar Crossref Search ADS PubMed 24 WHO . Global tuberculosis control 2012. Geneva: World Health Organization, 2012 . 25 Dodd PJ , Gardiner E , Coghlan R , et al. Burden of childhood tuberculosis in 22 high-burden countries: a mathematical modelling study . Lancet Glob Health 2014 ; 2 : 453 – 9 . Google Scholar Crossref Search ADS 26 Miedema KG , De Bont ES , Oude Nijhuis CS , et al. Validation of a new risk assessment model for predicting adverse events in children with fever and chemotherapy-induced neutropenia . J Clin Oncol 2011 ; 29 : e182 – 4 . Google Scholar Crossref Search ADS PubMed 27 Gomez E , Agrawal AK , Hastings CA , Transfusion support. In: Supportive care in Pediatric Oncology: A Practical Evidence-based Approach, Pediatric Oncology . Berlin . Springer Heidelberg , 2015 . 28 Stones DK , De Bruin GP , Esterhuizen TM , et al. Childhood cancer survival rates in two South African units . S Afr Med J 2014 ; 104 : 502 – 4 . Google Scholar Crossref Search ADS 29 Secmeer G , Devrim I , Kara A , et al. Role of procalcitonin and CRP in differentiating a stable from a deteriorating clinical course in pediatric febrile neutropenia . J Pediatr Hematol Oncol 2007 ; 29 : 107 – 11 . Google Scholar Crossref Search ADS PubMed 30 Miedema KG , De Bont ES , Elferink RF , et al. The diagnostic value of CRP, IL-8, PCT, and sTREM-1 in the detection of bacterial infections in pediatric oncology patients with febrile neutropenia . Support Care Cancer 2011 ; 19 : 1593 – 600 . Google Scholar Crossref Search ADS PubMed 31 Kitanovski L , Jazbec J , Hojker S , et al. Diagnostic accuracy of procalcitonin and interleukin-6 values for predicting bacteremia and clinical sepsis in febrile neutropenic children with cancer . Eur J Clin Microbiol Infect Dis 2006 ; 25 : 413 – 15 . Google Scholar Crossref Search ADS PubMed 32 Mian A , Becton D , Saylors R , et al. Biomarkers for risk stratification of febrile neutropenia among children with malignancy: a pilot study . Pediatr Blood Cancer 2012 ; 59 : 238 – 45 . Google Scholar Crossref Search ADS PubMed © The Author [2017]. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

Journal of Tropical Pediatrics – Oxford University Press

Published: Oct 1, 2018

There are no references for this article.